1. Field of the Invention
The invention relates to gas separation processes and more particularly the separation and recovery (or disposal) of perfluorocompound gases from a gas mixture. Especially, the invention relates to the concentrating of low concentration gas mixtures of perfluorocompound gases such as those present in the effluent of a semiconductor manufacturing process, particularly the etching and cleaning steps.
2. Related Art
The semiconductor industry is now using extensively perfluorocompounds such as CF.sub.4, C.sub.2 F.sub.6, C.sub.3 F.sub.8, C.sub.4 F.sub.10, CHF.sub.3, SF.sub.6, NF.sub.3, and the like, in the semiconductor manufacturing processes involving gases, particularly in the various etching steps of the semiconductor manufacturing processes as well as in the chamber cleaning step of the manufacturing process. Those perfluorocompound gases are used either pure or diluted, for example with air or nitrogen or other inert gas or in admixture with other perfluorocompound gases or other carrier gases (for example inert gases). All of those gases do not necessarily react with other species during the manufacturing processes: accordingly, when reactors are cleaned or evacuated to carry out another step of the manufacturing process, the effluent gases or gas mixtures should not be vented, even if they are largely diluted with air or any other gas such as inert gas. Most of the perfluorocompounds (also called PFCs) have lifetimes measured in thousands of years in the atmosphere and are also strong infrared absorbers. In the "Global Warming Symposium" held on Jun. 7-8, 1994, in Dallas, Tex., USA, carbon tetrafluoride (CF.sub.4), hexafluoroethane (C.sub.2 F.sub.6), nitrogen trifluoride (NF.sub.3), and sulfur hexafluoride (SF.sub.6) have been identified as greenhouse gases of concern to the semiconductor industry.
In the presentation made at this symposium by Michael T. Mocella and entitled "Perfluorocompound Emission Reduction From Semiconductor Processing Tools: An Overview Of Options And Strategies", the various possible strategies to control emission of these gases in the atmosphere were explained.
Apart from process replacement by non PFCs, several methods are already known or under development:
chemical-thermal decomposition with various activated metals wherein the spent bed material must be disposed. It is presently considered as commercially unproven even if it is under promising development. PA0 combustion-based decomposition process (or chemical-thermal process) using a flame to supply both the thermal energy and the reactants for the decomposition. There are some safety issues associated with the hydrogen or natural gas fuels used and all the PFCs will produce HF as a combustion product (if the temperature is high enough), whose emissions are also of concern and must be dealt with also. High temperatures may also be generated using a resistance heater. PA0 plasma-based decomposition process which involves the use of a plasma such as coupled radio frequency systems to partially decompose C.sub.2 F.sub.6, with over 90% decomposition of C.sub.2 F.sub.6. However, such systems are not yet commercially proven. Oxygen is usually needed to drive the decomposition to non PFC products with, however, the generation of HF which needs to be thereafter managed. PA0 recovery process wherein the PFCs are recovered instead of being destroyed in order to be recycled. This kind of process is of a great interest because it is considered as the "greenest" one. Different schemes, according to the author, are possible "based on combinations of adsorption or low temperature trapping of PFCs". There are, however, several challenges such as dealing with the large amount of nitrogen associated with the pump operation, the close boiling points of CF.sub.4 and NF.sub.3, the mixing of various process streams and/or possible reactions with adsorbents. While recycle is suggested, there are obvious questions about recycling such mixtures.
Another combustion system for destroying high nitrogen content effluent gas streams comprising PFCs is disclosed in the article entitled "Vector Technology's Phoenix Combustor" by Larry Anderson presented at the same symposium Jun. 7-8, 1994. This abatement system also uses a gas flame (using natural gas or hydrogen with air), which leads then to the same problem of HF generation and further destruction (plus the generation of NO.sub.x, CO.sub.2 inherent to any combustion process).
In the article presented at the same symposium by AT&T Microelectronics and Novapure Corporation and entitled "PFC Concentration and Recycle", the authors acknowledge the advantages of the recovery processes which avoid production of carbon dioxide, NO.sub.x and HF (compared to combustion processes). Briefly, this process is disclosed as the use of a dual bed adsorber (activated carbon), wherein one of the beds is in the adsorption mode, while the second bed is regenerated: the PFCs are adsorbed on the carbon sieves while the "carrier" gases, such as nitrogen, hydrogen are not adsorbed and are vented to the exhaust system. When the system is switched on the second adsorber, then the first one is evacuated using a vacuum pump, then the effluent is recompressed and the PFC gas mixture is recovered. One of the issues not yet resolved with such a system is that CF.sub.4, which is non polar, is not readily adsorbed by the carbon sieve and is then rejected with the vent gases. Also, any adsorption system is very sensitive to moisture and any trace of water has to be removed from the feed flow.
It is also known from U.S. Pat. No. 5,281,255 to use membranes made of rubbery polymers such as poly dimethyl siloxane or certain particular polymers such as a substituted polyacethylene (having a low glass transition temperature), to recover condensable organic components having a boiling point higher than -50.degree. C., essentially hydrocarbons (CH.sub.4, C.sub.2 H.sub.6, and the like), said hydrocarbons having the property of permeating through said membranes much faster than air, and then recovering on the permeate side of the membrane said hydrocarbons. The permeate (hydrocarbons) is then recovered at either substantially atmospheric pressure or lower pressure while the non-permeate (e.g. air) is still at the original pressure of the feed stream but is vented, and all of the pressure energy of the feed stream is lost.
Also, it is disclosed in U.S. Pat. No. 5,051,114, a selectively permeable membrane formed from an amorphous polymer of perfluoro 2-2 dimethyl 1-3-dioxole which is usable for separation of hydrocarbons or chlorofluorocarbons from, for example, air. Such a particular membrane apparently permeates oxygen and nitrogen faster than hydrocarbons and chlorofluorocarbons which can be recovered unexpectedly on the non-permeate side of the membrane, contrary to all of the membranes, including those disclosed in U.S. Pat. Nos. 4,553,983 and 5,281,255. In the '114 patent, there is also disclosed a mixture of the amorphous polymer of perfluoro 2-2 dimethyl 1-3 dioxole and polytetrafluoroethylene. All these perfluoro polymers are known to be resistant to most of the harmful chlorofluorocarbons and hydrocarbons which make them commercially suitable for such separation. However, this membrane is not currently available and there is no indication in this patent whether or not such a membrane is suitable for separation of PFCs from air or nitrogen particular at low concentrations of PFCs in carrier gases, and at widely varying feed flow conditions.
There is still presently a need for a "green" process for concentration and/or recovery of PFCs from a gaseous stream, which can be used with a feed flow comprising or saturated with, moisture, which can handle safely the PFCs recovery and/or concentration even with important or extreme variations of flows and/or concentration of PFCs in the feed stream which does not produce hydrofluoric acid (HF) as a residue from the destruction of the PFCs (in addition to the possible HF content of the feed).